- Email: [email protected]
Is the visibility of standardized inflicted bruises improved by using an alternate (‘forensic’) light source?
Accepted Manuscript Title: Is the visibility of standardized inflicted bruises improved by using an alternate (‘forensic’) light source? Authors: H.G.T. Nijs, R. De Groot, M.F.A.M. Van Velthoven, R.D. Stoel PII: DOI: Reference:
S0379-0738(18)30998-8 https://doi.org/10.1016/j.forsciint.2018.10.029 FSI 9532
To appear in:
FSI
Received date: Revised date: Accepted date:
13 June 2018 26 October 2018 29 October 2018
Please cite this article as: H.G.T.Nijs, R.De Groot, M.F.A.M.Van Velthoven, R.D.Stoel, Is the visibility of standardized inflicted bruises improved by using an alternate (‘forensic’) light source?, Forensic Science International https://doi.org/10.1016/j.forsciint.2018.10.029 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Is the visibility of standardized inflicted bruises improved by using an alternate (‘forensic’) light source? H.G.T. Nijs MD PhD, R. De Groot, M.F.A.M. Van Velthoven, R.D. Stoel PhD Netherlands Forensic Institute, Laan van Ypenburg 6, 2497 GB The Hague, The Netherlands Corresponding author: [email protected]
IP T
Other authors: [email protected] [email protected]
SC R
[email protected] Highlights (revision 2)
A weight was dropped onto the forearm of volunteers to create bruises
Bruises were photographed using white and alternate light at 415 nm
Photographs were evaluated by 10 assessors independently
Bruises were slightly better visible with alternate light at days 1 and 2
No differences were found in bruises aged 0.25, 7 and 14 days
A
N
U
M
Abstract Aim: to study the visibility of standardized inflicted bruises by using an alternate
ED
(‘forensic’) light source compared to a white light source. Methods: bruises were inflicted on the flexor site of the forearm (halfway in the middle) in 76 adults, by suddenly allowing a cylindrical metal object (400 grams) with rounded edges
PT
to drop for 1 meter in a vertically positioned tube. At 0.25, 1, 2, 7 and 14 days after this blunt force impact, the impact site on the forearm was photographed with a white light
CC E
source and subsequently with an alternate light source at 415 nm. Visibility of bruises on 170 randomized photographs was assessed on a calibrated monitor by 10 forensic medical specialists (physicians and pathologists) independently in two sessions: 1) with white light source photographs, and 2) after a mean of 11 days with greyscale converted alternate light source photographs.
A
Bruise visibility was expressed as a report mark between 1 (very bad) and 10 (excellent), or as ‘no visible bruise’. To determine intra-rater agreement, 10 of 170 photographs were assessed twice (untold to the assessors). In total 3600 (180*10*2) photographs were assessed. Results: 39 of 73 (53%) participants who completed the study, developed a visible bruise (women more often than men, p<0.001). Inter-rater agreement between assessors was high (mean inter-class coefficient, ICC, for white light source 0.66 (SD 0.14) and for
alternate light source ICC 0.73 (0.09)). Intra-rater agreement was excellent (mean ICC 0.88 (SD 0.09)). Mean report marks for bruise visibility, recorded independently by 10 assessors on 170 unique photographs per light source, were significantly higher with an alternate light source than with a white light source, at 1 and 2 days after impact: 4.4 (SD 2.0) vs 3.8 (1.8) (p<0.01) and 4.9 (2.1) vs 4.5 (2.0) (p<0.05), respectively. However, these differences were small, as the mean difference (effect size) in report marks were 0.6 (0.5) and 0.4 (0.3), at 1 and 2 days after impact, respectively.
IP T
The other time points showed no statistical significant differences in report marks. Conclusions: bruises after standardized blunt force impact were slightly better visible with
an alternate light source than with a white light source after 1 and 2 days, but not after 0.25, 7 and 14 days. The value of using an alternate light source at 415 nm to improve
SC R
bruise visibility was limited in this study.
Keywords: alternate light source
U
forensic light source blunt force trauma
N
bruising
M
A
forensic medicine
Introduction
ED
Bruises are typically caused by blunt force trauma to the skin [1]. By damaging underlying blood vessels or capillaries while the skin remains intact, blood extravasates into surrounding tissues. After some variable time a (sub)cutaneous blood/fluid collection (hematoma, bruise) may become visible to the naked eye; this usually discolorates and
PT
disappears within 2-3 weeks. [2]
Medical forensic evaluation of bruises is important in children and adults, alive or
CC E
deceased, to differentiate accidental from non-accidental causes in relation to possible medical conditions [3] [4]. Many investigators including technical police officers believe that bruise visibility will be improved by enhancing contrast by using an alternate light source compared to a white light source or available light. However, the added value of an alternate (‘forensic’) light source has actually little been studied in vivo.
A
One study showed that an alternate light source has limited increased sensitivity compared to the naked eye in detecting known bruises, except after 14 days [5]. In children limited improved visibility of bruises was found by using cross-polarized light [6]. Since peak absorption for hemoglobin is at 415 nm, alternate light source at this wavelength seems the best way to visualize bruises [7]. The aim of our study was to compare the visibility of standardized inflicted bruises in healthy adults by using an alternate light source at 415 nm and a white light source.
A ED
PT
CC E
IP T
SC R
U
N
A
M
Methods Recruitment of subjects Volunteers were recruited from employees of the Netherlands Forensic Institute (NFI), The Hague, The Netherlands, with leaflets on frequently visited areas (e.g. lunch room) and two general messages on the Intranet with a description of the study. Any healthy Caucasian adult, 18-65 years of age regardless of gender, race or socioeconomic group was eligible. Exclusion criteria included any acute systematic illness, and a medical condition or medication that could possibly affect coagulation or bruising.
IP T
One of the researchers (MVV) approached potential subjects, made the appointment, explained the study, answered questions and obtained written informed consent.
Seventy six volunteers entered the study. Participants filled in a questionnaire about gender, age, sport activities and prescribed medication. Perimeter and skin crease
SC R
thickness at the intended impact site of the forearm (see below) were measured with a centimeter and a caliper. Standardized bruise infliction
U
We used the non-dominant forearm, for easy and comfortable access. One of the researchers (MVV) inspected the skin of the intended impact site with white and alternate
N
light sources, for any pre-existing skin discolorations or abnormalities. No participants were excluded based on this inspection.
A
Bruises were inflicted on the flexor site of the forearm, halfway between wrist and elbow
M
pleat and in the middle between the ulnar and radial side of the forearm (‘halfway in the middle’). In this way the point of impact could easily be retraced, even weeks later. By sudden unexpected release of a pin (top arrow Figure 1A), a cylindrical metal object
ED
(400 grams) with rounded edges was allowed to drop for 1 meter in a vertically positioned tube (bottom arrow Figure 1A), in order to produce a standardized blunt force impact (method according to Lombardi et al. [5], with slightly modified object weight and fall
PT
height). Directly following impact, participants were asked to express perceived pain as a report mark ranging from 1 (no pain whatsoever) to 10 (extreme pain) (=pain score).
CC E
Photography of impact site The impact site on the forearm was photographed under standard conditions with a (photographic) white light source, followed by an alternate light source (Projectina, SL450, Switzerland) at 415 nm, both light sources at fixed distances (see Figure 1B), at
A
0.25, 1, 2, 7 and 14 days after impact (the last time point only if at 7 days a bruise was visible with either light source). When using the alternate light source, the room was darkened and all individuals wore protection goggles. A fixed camera (Nikon D3) was used (see Figure 1B, middle), perpendicular to the surface, with a 60 mm macro lens and a preset white balance. For the white light source, aperture F9, shutter speed 1/80 sec and ISO 400 were used. For the alternate light source, aperture F9, shutter speed 1 second, ISO 4.000 and filter 023 (yellow) were used.
DNG-files were processed in Lightroom® 6.0 and converted to .JPG-files. To maximize contrast for photographs with the alternate light source, greyscale conversion was applied using Lightroom (inspired by Olds et al) [8]. Two typical examples of photographs, as
N
U
SC R
IP T
displayed to the assessors, are shown in Figures 2A and 2B.
Figure 1B
A
Figure 1A
Figure 1A Standardized bruise infliction (see Text) (photograph published with permission
M
of the male volunteer not participating in the study and the female investigator MVV). Figure 1B Test setup for photography of impact sites with a white light source (left and
A
CC E
PT
ED
right) and an alternate light source (middle/right) with a fixed camera (middle) (see Text).
Figure 2A
Figure 2B
Figure 2A Typical example of photograph with a white light source. Figure 2B Typical example of a greyscale converted photograph with an alternate light source (same forearm and same time point, 2 days, as in Figure 2A).
This part of the study was performed from October 2016 to March 2017. The study was approved by the (external) Medical Research Ethics Committee of Academic Medical Center, Amsterdam, The Netherlands (NL 53398.018.15). Assessment of bruise visibility We designed a method to assess bruise visibility on photographs. Photographs were randomized by time point after impact and by participant; the same order was used for each assessor. To reduce work load, photographs of 34 participants were excluded in case two young researchers (MVV and RDG) saw no visible bruise at any time with either light source. In total 170 unique photographs of 39 participants, who showed a bruise at any
IP T
time during the study, were displayed for assessment (e.g. Figure 1C and 1D) by using an
online survey (Survey Monkey®). To determine intra-rater agreement, 10 out of these 170
SC R
photographs were displayed twice dispersed through the survey, untold to the assessors.
Photographs were independently assessed on a single calibrated monitor by a total of 10 forensic medical specialists (physicians and pathologists) of our institution. In the first session, 180 (170 + 10 double) photographs with white light source were displayed. In the second session after a mean of 11 days (range 1-31 days), the same series of 180
U
photographs, but then with an alternate light source and greyscale converted, was
N
displayed. In total 3600 photographs were assessed (180 photographs * 10 assessors * 2
A
sessions) from January to April 2018.
Bruise visibility was expressed by the assessor as a report mark between 1 (very bad) and
M
10 (excellent) according to the Dutch report grade system, well-known to the assessors, or
Statistical analysis
ED
as ‘no visible bruise’ (according to the assessor).
Results were analyzed with SPSS® 24. Mean report marks were compared using t-tests.
PT
Intra-rater and inter-rater agreement were analyzed by using intra-class correlation coefficients (ICC). Subjects’ and other characteristics were analyzed by using t-test and
A
CC E
logistic regression. A p-value of 0.05 or less was considered as statistically significant.
Results Demographics In total 76 volunteers entered the study. Three participants were lost to follow-up (defined as 3 or more missed visits out of the 5 visits). Hence, 73 participants remained for data analysis (36 males, 37 females, mean age 37 (SD 11.8) years). Of 73 participants, 39 (53.4%) developed a visible bruise at any time during the study. Of 34 participants who did not develop a visible bruise, 76.5% were male (p<0.05 compared to females). Logistic regression analysis with pain score, skin crease thickness developing a bruise (female versus male, odds ratio (OR) 7.7, p<0.001).
IP T
and forearm perimeter as co-variables showed female sex to be positively related in The mean pain score after impact was 4.5 (SD 1.8). A higher pain score was associated with more visible bruises (for high pain score versus low pain score, median split, OR 1.5,
SC R
p<0.05). Inter- and intra-rater agreement
Inter-rater agreement was considered high (mean inter-class coefficient, ICC 0.66 (SD 0.14) for white light source, and ICC 0.73 (0.09) for alternate light source).
N
U
Intra-rater agreement was considered excellent (mean ICC 0.88 (SD 0.09)). Visibility of bruises
A
According to the report marks recorded by the assessors, most of the bruises were visible both with an alternate light source as well as with a white light source. The score ‘no
M
visible bruise’ was given more often with an alternate light source than with a white light source (respectively 379 versus 319 out of 1800, p< 0.02).
ED
Mean report marks for bruise visibility, as recorded independently by 10 assessors on 170 unique photographs per light source, are show in Table 1. Mean report marks for bruise visibility with an alternate light source compared to a white light source were significantly higher at 1 and 2 days after impact: 4.4 (SD 2.0) vs 3.8 (1.8) (p<0.01) and 4.9 (2.1) vs
PT
4.5 (2.0) (p<0.05), respectively.
However, these differences were small, as the mean difference (effect size) was 0.6 (0.5)
CC E
and 0.4 (0.3) at 1 and 2 days after impact, respectively. The other time points (0.25, 7 and 14 days) showed no statistical significant differences in
A
report marks for bruise visibility.
N
Report mark white light Mean (SD)
Report mark alternate light Mean (SD)
Report mark Mean (SD) difference1
Report mark Effect size2 of difference
p-value3
39
2.4 (1.2)
2.1 (1.1)
-0.2 (1.2)
-0.2
0.23
1
37
3.8 (1.8)
4.4 (2.0)
+0.6 (1.3)
+0.5
<0.01
2
39
4.5 (2.0)
4.9 (2.1)
+0.4 (1.2)
+0.3
0.04
7
34
3.4 (2.1)
3.1 (2.1)
-0.3 (1.0)
-0.3
0.13
14
21
2.1 (1.4)
2.3 (1.1)
+0.3 (1.1)
+0.3
0.34
IP T
Time after impact (days) 0.25
A
CC E
PT
ED
M
A
N
U
SC R
Table 1 Report marks, as recorded independently by 10 assessors on 170 unique photographs per light source, ranging from 1 (very bad) to 10 (excellent) for bruise visibility with a white light source and with an alternate light source, at different time intervals after standardized bruise infliction (see Text). 1 mean report mark of alternate light source minus white light source 2 effect size= mean difference/SD 3 two-sided p-value
Discussion To study the visibility of bruises by using an alternate light source at 415 nm compared to a white light source, we inflicted standardized bruises on a forearm in 76 healthy adults. The mean report mark for bruise visibility, independently assessed on randomized photographs by 10 forensic medical specialists (physicians and pathologists) of our institution, was slightly higher (i.e. better visibility) with an alternate light source than with a white light source at 1 and 2 days after standardized blunt force impact. The other time points (0.25, 7 and 14 days after impact) showed no statistical differences in report marks
IP T
for bruise visibility.
Our results, showing limited value of an alternate light source compared to a white light
source to visualize bruises, are in line with the few other available studies. Lombardi et al,
SC R
using a similar method of bruise infliction (with slightly different object weight and fall height), found limited increased sensitivity to detect subclinical bruising for an alternate light source except after 14 days (on days 1, 7 and 14: 76.8%, 69.6% and 60.7% for alternate light source versus 69.6%, 60.0% and 32.1% for white light) [5]. Limited
U
improved visibility was found by using cross-polarized light in children [6], as well as in a pigskin model with injection of blood [8] and in human cadavers [9].
N
Also, increased visibility of bruises was found in survivors of strangulation [10] or other
A
various types of violence [11]. However, these studies were not standardized with respect to the bruises. We are unaware of other published studies addressing this subject in adults
M
in vivo with standardized inflicted bruises, like we did.
When taking subject characteristics into account, we found that women developed more
ED
often a bruise than men. We hypothesized this could be due to the larger amount of subcutaneous fat in women compared to men. To explore this issue we used logistic regression models with pain score, skin crease and forearm perimeter as co-variables, but
PT
found no differences by doing so. Further research is required. The limitations of our study follow from the standardization of the study: we only used one
CC E
method to inflict standardized bruises at one location in heathy Caucasian adults (white skin) and of course not in children, who may have different bruise characteristics. Moreover, only one (yet often used) wavelength, 415 nm, was studied.
A
We converted photographs with alternate light source to greyscale in order to optimize contrast with surrounding skin (inspired by Olds et al [8]). The results of our study are based on these converted photographs. Different conversion settings were not investigated for practical reasons. The strengths of our study are standardization of the method of bruise infliction with a plausible bruise mechanism, and that we studied healthy adults in vivo.
We also are confident that the method used to assess bruise visibility on photographs, by using a fairly large number (10) of forensic medical specialists (physicians and pathologists), who were unaware of the time point and subject, is a nice approximation as ‘gold standard’. Inter- and intra-rater agreement was high. With the method described we found visible bruises in 53% of participants at acceptable pain scores (mean 4.5 out of 10). From personal experience we know that a more highenergetic impact, for example by increasing object weight or fall height, is not acceptable.
IP T
We did not perform paired comparisons in the assessment of bruise visibility, as might be performed in reality by assessing photographs with white light source and alternate light source simultaneously or by flipping these modalities forwards and backwards. However, in
a preliminary study with simultaneous paired comparisons by one researcher similar
SC R
results were found as in the present study [12].
We were surprised by the absence of differences in bruise visibility between the two light sources, especially in ‘the long run’, 7 and 14 days after impact. Because it can be
U
hypothesized (and in our experience is often assumed by forensic investigators and technical police officers) that in case of ‘chronic injury’ an alternate light source may be
N
better in visualizing bruises than a white light source. This was not the case in our study (in contrast to Lombardi, who found substantially increased sensitivity of bruise detection
A
after 14 days, but not after 7 days, with alternate light source compared to white light)
M
[10].
In conclusion, bruises after standardized blunt force impact were slightly better visible with
ED
an alternate light source than with a white light source after 1 and 2 days, but not after 0.25, 7 and 14 days. The value of using an alternate light source at 415 nm to improve
PT
bruise visibility was limited in this study. Author contribution
CC E
H.G.T. Nijs: study design, analysis, reporting R. De Groot: performing the study, analysis, reporting M.F.A.M. van Velthoven: performing the study, reporting
A
R.D.Stoel: study design, analysis, reporting Acknowledgments We would like to express our thanks to our colleagues: as participants in the study or as assessors of a large number of photographs.
References
[1]
N. E. I. Langlois, “The science behind the quest to determine the age of bruises - a review of the English language literature,” Forensic Sci. Med. Pathol., vol. 3, pp. 241–251, 2007.
[2]
R. A. C. Bilo, A. P. Oranje, T. Shwayder, and C. J. Hobbs, Cutaneous Manifestations of Child Abuse and Their Differential Diagnosis: Blunt Force Trauma. Springer-Verlag Berlin Heidelberg, 2013.
[3]
F. Fersini, A. Govi, M. Tsokos, S. Etzold, and L. Tattoli, “Examples of
IP T
tramline bruises in clinical forensic medicine,” Forensic Sci. Med. Pathol., vol. 13, pp. 508–510, 2017. [4]
P. Vanezis, “Interpreting bruises at necropsy,” J. Clin. Pathol., vol. 54, pp.
[5]
SC R
348–355, 2001.
M. Lombardi, J. Canter, P. A. Patrick, and R. Altman, “Is Fluorescence Under an
Alternate
Light
Source Sufficient
to Accurately Diagnose
Subclinical Bruising?,” J. Forensic Sci., vol. 60, no. 2, pp. 444–449, 2015. L. Trefan, C. Harris, S. Evans, D. Nuttall, S. Maguire, and A. M. Kemp, “A
U
[6]
comparison of four different imaging modalities - conventional, cross
N
polarized, infra-red and ultra-violet in the assessment of childhood [7]
A
bruising,” J. Forensic Leg. Med., vol. 59, pp. 30–35, 2018. V. K. Hughes, P. S. Ellis, and N. E. I. Langlois, “Alternative light source
M
(polilight®) illumination with digital image analysis does not assist in determining the age of bruises,” Forensic Sci. Int., vol. 158, no. 2–3, pp. [8]
ED
104–107, 2006.
K. Olds, R. W. Byard, C. Winskog, and N. E. I. Langlois, “Validation of ultraviolet , infrared , and narrow band light alternate light sources for
PT
detection of bruises in a pigskin model,” Forensic Sci. Med. Pathol., vol. 12, no. 4, pp. 435–443, 2016. K. Olds, R.W. Byard, C. Winskog, N.E. Langlois. Validation of alternate light
CC E
[9]
sources for detection of bruises in non-embalmed and embalmed cadavers. Forensic Sci. Med. Pathol., vol. 13, no. 1, pp. 28-33, 2017.
[10]
D.S. Holbrook, M.C. Jackson. Use of an alternative light source to assess
A
strangulation victims. J Forensic Nurs., vol. 9, no. 3, pp. 140-5, 2013.
[11]
R.M. Limmen, M. Ceelen, U.J.L. Reijnders, J. Stomp et al. Enhancing the Visibility of Injuries with Narrow-Banded Beams of Light within the Visible Light Spectrum. J Forensic Sci., vol. 58, pp. 518-22, 2013.
[12]
M. F. A. M. van Velthoven and H. G. T. Nijs, “Is the visibility of standardized induced bruises improved under a forensic light source?,” in EAFS, Prague, Abstract O211, 2015.
S0379-0738(18)30998-8 https://doi.org/10.1016/j.forsciint.2018.10.029 FSI 9532
To appear in:
FSI
Received date: Revised date: Accepted date:
13 June 2018 26 October 2018 29 October 2018
Please cite this article as: H.G.T.Nijs, R.De Groot, M.F.A.M.Van Velthoven, R.D.Stoel, Is the visibility of standardized inflicted bruises improved by using an alternate (‘forensic’) light source?, Forensic Science International https://doi.org/10.1016/j.forsciint.2018.10.029 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Is the visibility of standardized inflicted bruises improved by using an alternate (‘forensic’) light source? H.G.T. Nijs MD PhD, R. De Groot, M.F.A.M. Van Velthoven, R.D. Stoel PhD Netherlands Forensic Institute, Laan van Ypenburg 6, 2497 GB The Hague, The Netherlands Corresponding author: [email protected]
IP T
Other authors: [email protected] [email protected]
SC R
[email protected] Highlights (revision 2)
A weight was dropped onto the forearm of volunteers to create bruises
Bruises were photographed using white and alternate light at 415 nm
Photographs were evaluated by 10 assessors independently
Bruises were slightly better visible with alternate light at days 1 and 2
No differences were found in bruises aged 0.25, 7 and 14 days
A
N
U
M
Abstract Aim: to study the visibility of standardized inflicted bruises by using an alternate
ED
(‘forensic’) light source compared to a white light source. Methods: bruises were inflicted on the flexor site of the forearm (halfway in the middle) in 76 adults, by suddenly allowing a cylindrical metal object (400 grams) with rounded edges
PT
to drop for 1 meter in a vertically positioned tube. At 0.25, 1, 2, 7 and 14 days after this blunt force impact, the impact site on the forearm was photographed with a white light
CC E
source and subsequently with an alternate light source at 415 nm. Visibility of bruises on 170 randomized photographs was assessed on a calibrated monitor by 10 forensic medical specialists (physicians and pathologists) independently in two sessions: 1) with white light source photographs, and 2) after a mean of 11 days with greyscale converted alternate light source photographs.
A
Bruise visibility was expressed as a report mark between 1 (very bad) and 10 (excellent), or as ‘no visible bruise’. To determine intra-rater agreement, 10 of 170 photographs were assessed twice (untold to the assessors). In total 3600 (180*10*2) photographs were assessed. Results: 39 of 73 (53%) participants who completed the study, developed a visible bruise (women more often than men, p<0.001). Inter-rater agreement between assessors was high (mean inter-class coefficient, ICC, for white light source 0.66 (SD 0.14) and for
alternate light source ICC 0.73 (0.09)). Intra-rater agreement was excellent (mean ICC 0.88 (SD 0.09)). Mean report marks for bruise visibility, recorded independently by 10 assessors on 170 unique photographs per light source, were significantly higher with an alternate light source than with a white light source, at 1 and 2 days after impact: 4.4 (SD 2.0) vs 3.8 (1.8) (p<0.01) and 4.9 (2.1) vs 4.5 (2.0) (p<0.05), respectively. However, these differences were small, as the mean difference (effect size) in report marks were 0.6 (0.5) and 0.4 (0.3), at 1 and 2 days after impact, respectively.
IP T
The other time points showed no statistical significant differences in report marks. Conclusions: bruises after standardized blunt force impact were slightly better visible with
an alternate light source than with a white light source after 1 and 2 days, but not after 0.25, 7 and 14 days. The value of using an alternate light source at 415 nm to improve
SC R
bruise visibility was limited in this study.
Keywords: alternate light source
U
forensic light source blunt force trauma
N
bruising
M
A
forensic medicine
Introduction
ED
Bruises are typically caused by blunt force trauma to the skin [1]. By damaging underlying blood vessels or capillaries while the skin remains intact, blood extravasates into surrounding tissues. After some variable time a (sub)cutaneous blood/fluid collection (hematoma, bruise) may become visible to the naked eye; this usually discolorates and
PT
disappears within 2-3 weeks. [2]
Medical forensic evaluation of bruises is important in children and adults, alive or
CC E
deceased, to differentiate accidental from non-accidental causes in relation to possible medical conditions [3] [4]. Many investigators including technical police officers believe that bruise visibility will be improved by enhancing contrast by using an alternate light source compared to a white light source or available light. However, the added value of an alternate (‘forensic’) light source has actually little been studied in vivo.
A
One study showed that an alternate light source has limited increased sensitivity compared to the naked eye in detecting known bruises, except after 14 days [5]. In children limited improved visibility of bruises was found by using cross-polarized light [6]. Since peak absorption for hemoglobin is at 415 nm, alternate light source at this wavelength seems the best way to visualize bruises [7]. The aim of our study was to compare the visibility of standardized inflicted bruises in healthy adults by using an alternate light source at 415 nm and a white light source.
A ED
PT
CC E
IP T
SC R
U
N
A
M
Methods Recruitment of subjects Volunteers were recruited from employees of the Netherlands Forensic Institute (NFI), The Hague, The Netherlands, with leaflets on frequently visited areas (e.g. lunch room) and two general messages on the Intranet with a description of the study. Any healthy Caucasian adult, 18-65 years of age regardless of gender, race or socioeconomic group was eligible. Exclusion criteria included any acute systematic illness, and a medical condition or medication that could possibly affect coagulation or bruising.
IP T
One of the researchers (MVV) approached potential subjects, made the appointment, explained the study, answered questions and obtained written informed consent.
Seventy six volunteers entered the study. Participants filled in a questionnaire about gender, age, sport activities and prescribed medication. Perimeter and skin crease
SC R
thickness at the intended impact site of the forearm (see below) were measured with a centimeter and a caliper. Standardized bruise infliction
U
We used the non-dominant forearm, for easy and comfortable access. One of the researchers (MVV) inspected the skin of the intended impact site with white and alternate
N
light sources, for any pre-existing skin discolorations or abnormalities. No participants were excluded based on this inspection.
A
Bruises were inflicted on the flexor site of the forearm, halfway between wrist and elbow
M
pleat and in the middle between the ulnar and radial side of the forearm (‘halfway in the middle’). In this way the point of impact could easily be retraced, even weeks later. By sudden unexpected release of a pin (top arrow Figure 1A), a cylindrical metal object
ED
(400 grams) with rounded edges was allowed to drop for 1 meter in a vertically positioned tube (bottom arrow Figure 1A), in order to produce a standardized blunt force impact (method according to Lombardi et al. [5], with slightly modified object weight and fall
PT
height). Directly following impact, participants were asked to express perceived pain as a report mark ranging from 1 (no pain whatsoever) to 10 (extreme pain) (=pain score).
CC E
Photography of impact site The impact site on the forearm was photographed under standard conditions with a (photographic) white light source, followed by an alternate light source (Projectina, SL450, Switzerland) at 415 nm, both light sources at fixed distances (see Figure 1B), at
A
0.25, 1, 2, 7 and 14 days after impact (the last time point only if at 7 days a bruise was visible with either light source). When using the alternate light source, the room was darkened and all individuals wore protection goggles. A fixed camera (Nikon D3) was used (see Figure 1B, middle), perpendicular to the surface, with a 60 mm macro lens and a preset white balance. For the white light source, aperture F9, shutter speed 1/80 sec and ISO 400 were used. For the alternate light source, aperture F9, shutter speed 1 second, ISO 4.000 and filter 023 (yellow) were used.
DNG-files were processed in Lightroom® 6.0 and converted to .JPG-files. To maximize contrast for photographs with the alternate light source, greyscale conversion was applied using Lightroom (inspired by Olds et al) [8]. Two typical examples of photographs, as
N
U
SC R
IP T
displayed to the assessors, are shown in Figures 2A and 2B.
Figure 1B
A
Figure 1A
Figure 1A Standardized bruise infliction (see Text) (photograph published with permission
M
of the male volunteer not participating in the study and the female investigator MVV). Figure 1B Test setup for photography of impact sites with a white light source (left and
A
CC E
PT
ED
right) and an alternate light source (middle/right) with a fixed camera (middle) (see Text).
Figure 2A
Figure 2B
Figure 2A Typical example of photograph with a white light source. Figure 2B Typical example of a greyscale converted photograph with an alternate light source (same forearm and same time point, 2 days, as in Figure 2A).
This part of the study was performed from October 2016 to March 2017. The study was approved by the (external) Medical Research Ethics Committee of Academic Medical Center, Amsterdam, The Netherlands (NL 53398.018.15). Assessment of bruise visibility We designed a method to assess bruise visibility on photographs. Photographs were randomized by time point after impact and by participant; the same order was used for each assessor. To reduce work load, photographs of 34 participants were excluded in case two young researchers (MVV and RDG) saw no visible bruise at any time with either light source. In total 170 unique photographs of 39 participants, who showed a bruise at any
IP T
time during the study, were displayed for assessment (e.g. Figure 1C and 1D) by using an
online survey (Survey Monkey®). To determine intra-rater agreement, 10 out of these 170
SC R
photographs were displayed twice dispersed through the survey, untold to the assessors.
Photographs were independently assessed on a single calibrated monitor by a total of 10 forensic medical specialists (physicians and pathologists) of our institution. In the first session, 180 (170 + 10 double) photographs with white light source were displayed. In the second session after a mean of 11 days (range 1-31 days), the same series of 180
U
photographs, but then with an alternate light source and greyscale converted, was
N
displayed. In total 3600 photographs were assessed (180 photographs * 10 assessors * 2
A
sessions) from January to April 2018.
Bruise visibility was expressed by the assessor as a report mark between 1 (very bad) and
M
10 (excellent) according to the Dutch report grade system, well-known to the assessors, or
Statistical analysis
ED
as ‘no visible bruise’ (according to the assessor).
Results were analyzed with SPSS® 24. Mean report marks were compared using t-tests.
PT
Intra-rater and inter-rater agreement were analyzed by using intra-class correlation coefficients (ICC). Subjects’ and other characteristics were analyzed by using t-test and
A
CC E
logistic regression. A p-value of 0.05 or less was considered as statistically significant.
Results Demographics In total 76 volunteers entered the study. Three participants were lost to follow-up (defined as 3 or more missed visits out of the 5 visits). Hence, 73 participants remained for data analysis (36 males, 37 females, mean age 37 (SD 11.8) years). Of 73 participants, 39 (53.4%) developed a visible bruise at any time during the study. Of 34 participants who did not develop a visible bruise, 76.5% were male (p<0.05 compared to females). Logistic regression analysis with pain score, skin crease thickness developing a bruise (female versus male, odds ratio (OR) 7.7, p<0.001).
IP T
and forearm perimeter as co-variables showed female sex to be positively related in The mean pain score after impact was 4.5 (SD 1.8). A higher pain score was associated with more visible bruises (for high pain score versus low pain score, median split, OR 1.5,
SC R
p<0.05). Inter- and intra-rater agreement
Inter-rater agreement was considered high (mean inter-class coefficient, ICC 0.66 (SD 0.14) for white light source, and ICC 0.73 (0.09) for alternate light source).
N
U
Intra-rater agreement was considered excellent (mean ICC 0.88 (SD 0.09)). Visibility of bruises
A
According to the report marks recorded by the assessors, most of the bruises were visible both with an alternate light source as well as with a white light source. The score ‘no
M
visible bruise’ was given more often with an alternate light source than with a white light source (respectively 379 versus 319 out of 1800, p< 0.02).
ED
Mean report marks for bruise visibility, as recorded independently by 10 assessors on 170 unique photographs per light source, are show in Table 1. Mean report marks for bruise visibility with an alternate light source compared to a white light source were significantly higher at 1 and 2 days after impact: 4.4 (SD 2.0) vs 3.8 (1.8) (p<0.01) and 4.9 (2.1) vs
PT
4.5 (2.0) (p<0.05), respectively.
However, these differences were small, as the mean difference (effect size) was 0.6 (0.5)
CC E
and 0.4 (0.3) at 1 and 2 days after impact, respectively. The other time points (0.25, 7 and 14 days) showed no statistical significant differences in
A
report marks for bruise visibility.
N
Report mark white light Mean (SD)
Report mark alternate light Mean (SD)
Report mark Mean (SD) difference1
Report mark Effect size2 of difference
p-value3
39
2.4 (1.2)
2.1 (1.1)
-0.2 (1.2)
-0.2
0.23
1
37
3.8 (1.8)
4.4 (2.0)
+0.6 (1.3)
+0.5
<0.01
2
39
4.5 (2.0)
4.9 (2.1)
+0.4 (1.2)
+0.3
0.04
7
34
3.4 (2.1)
3.1 (2.1)
-0.3 (1.0)
-0.3
0.13
14
21
2.1 (1.4)
2.3 (1.1)
+0.3 (1.1)
+0.3
0.34
IP T
Time after impact (days) 0.25
A
CC E
PT
ED
M
A
N
U
SC R
Table 1 Report marks, as recorded independently by 10 assessors on 170 unique photographs per light source, ranging from 1 (very bad) to 10 (excellent) for bruise visibility with a white light source and with an alternate light source, at different time intervals after standardized bruise infliction (see Text). 1 mean report mark of alternate light source minus white light source 2 effect size= mean difference/SD 3 two-sided p-value
Discussion To study the visibility of bruises by using an alternate light source at 415 nm compared to a white light source, we inflicted standardized bruises on a forearm in 76 healthy adults. The mean report mark for bruise visibility, independently assessed on randomized photographs by 10 forensic medical specialists (physicians and pathologists) of our institution, was slightly higher (i.e. better visibility) with an alternate light source than with a white light source at 1 and 2 days after standardized blunt force impact. The other time points (0.25, 7 and 14 days after impact) showed no statistical differences in report marks
IP T
for bruise visibility.
Our results, showing limited value of an alternate light source compared to a white light
source to visualize bruises, are in line with the few other available studies. Lombardi et al,
SC R
using a similar method of bruise infliction (with slightly different object weight and fall height), found limited increased sensitivity to detect subclinical bruising for an alternate light source except after 14 days (on days 1, 7 and 14: 76.8%, 69.6% and 60.7% for alternate light source versus 69.6%, 60.0% and 32.1% for white light) [5]. Limited
U
improved visibility was found by using cross-polarized light in children [6], as well as in a pigskin model with injection of blood [8] and in human cadavers [9].
N
Also, increased visibility of bruises was found in survivors of strangulation [10] or other
A
various types of violence [11]. However, these studies were not standardized with respect to the bruises. We are unaware of other published studies addressing this subject in adults
M
in vivo with standardized inflicted bruises, like we did.
When taking subject characteristics into account, we found that women developed more
ED
often a bruise than men. We hypothesized this could be due to the larger amount of subcutaneous fat in women compared to men. To explore this issue we used logistic regression models with pain score, skin crease and forearm perimeter as co-variables, but
PT
found no differences by doing so. Further research is required. The limitations of our study follow from the standardization of the study: we only used one
CC E
method to inflict standardized bruises at one location in heathy Caucasian adults (white skin) and of course not in children, who may have different bruise characteristics. Moreover, only one (yet often used) wavelength, 415 nm, was studied.
A
We converted photographs with alternate light source to greyscale in order to optimize contrast with surrounding skin (inspired by Olds et al [8]). The results of our study are based on these converted photographs. Different conversion settings were not investigated for practical reasons. The strengths of our study are standardization of the method of bruise infliction with a plausible bruise mechanism, and that we studied healthy adults in vivo.
We also are confident that the method used to assess bruise visibility on photographs, by using a fairly large number (10) of forensic medical specialists (physicians and pathologists), who were unaware of the time point and subject, is a nice approximation as ‘gold standard’. Inter- and intra-rater agreement was high. With the method described we found visible bruises in 53% of participants at acceptable pain scores (mean 4.5 out of 10). From personal experience we know that a more highenergetic impact, for example by increasing object weight or fall height, is not acceptable.
IP T
We did not perform paired comparisons in the assessment of bruise visibility, as might be performed in reality by assessing photographs with white light source and alternate light source simultaneously or by flipping these modalities forwards and backwards. However, in
a preliminary study with simultaneous paired comparisons by one researcher similar
SC R
results were found as in the present study [12].
We were surprised by the absence of differences in bruise visibility between the two light sources, especially in ‘the long run’, 7 and 14 days after impact. Because it can be
U
hypothesized (and in our experience is often assumed by forensic investigators and technical police officers) that in case of ‘chronic injury’ an alternate light source may be
N
better in visualizing bruises than a white light source. This was not the case in our study (in contrast to Lombardi, who found substantially increased sensitivity of bruise detection
A
after 14 days, but not after 7 days, with alternate light source compared to white light)
M
[10].
In conclusion, bruises after standardized blunt force impact were slightly better visible with
ED
an alternate light source than with a white light source after 1 and 2 days, but not after 0.25, 7 and 14 days. The value of using an alternate light source at 415 nm to improve
PT
bruise visibility was limited in this study. Author contribution
CC E
H.G.T. Nijs: study design, analysis, reporting R. De Groot: performing the study, analysis, reporting M.F.A.M. van Velthoven: performing the study, reporting
A
R.D.Stoel: study design, analysis, reporting Acknowledgments We would like to express our thanks to our colleagues: as participants in the study or as assessors of a large number of photographs.
References
[1]
N. E. I. Langlois, “The science behind the quest to determine the age of bruises - a review of the English language literature,” Forensic Sci. Med. Pathol., vol. 3, pp. 241–251, 2007.
[2]
R. A. C. Bilo, A. P. Oranje, T. Shwayder, and C. J. Hobbs, Cutaneous Manifestations of Child Abuse and Their Differential Diagnosis: Blunt Force Trauma. Springer-Verlag Berlin Heidelberg, 2013.
[3]
F. Fersini, A. Govi, M. Tsokos, S. Etzold, and L. Tattoli, “Examples of
IP T
tramline bruises in clinical forensic medicine,” Forensic Sci. Med. Pathol., vol. 13, pp. 508–510, 2017. [4]
P. Vanezis, “Interpreting bruises at necropsy,” J. Clin. Pathol., vol. 54, pp.
[5]
SC R
348–355, 2001.
M. Lombardi, J. Canter, P. A. Patrick, and R. Altman, “Is Fluorescence Under an
Alternate
Light
Source Sufficient
to Accurately Diagnose
Subclinical Bruising?,” J. Forensic Sci., vol. 60, no. 2, pp. 444–449, 2015. L. Trefan, C. Harris, S. Evans, D. Nuttall, S. Maguire, and A. M. Kemp, “A
U
[6]
comparison of four different imaging modalities - conventional, cross
N
polarized, infra-red and ultra-violet in the assessment of childhood [7]
A
bruising,” J. Forensic Leg. Med., vol. 59, pp. 30–35, 2018. V. K. Hughes, P. S. Ellis, and N. E. I. Langlois, “Alternative light source
M
(polilight®) illumination with digital image analysis does not assist in determining the age of bruises,” Forensic Sci. Int., vol. 158, no. 2–3, pp. [8]
ED
104–107, 2006.
K. Olds, R. W. Byard, C. Winskog, and N. E. I. Langlois, “Validation of ultraviolet , infrared , and narrow band light alternate light sources for
PT
detection of bruises in a pigskin model,” Forensic Sci. Med. Pathol., vol. 12, no. 4, pp. 435–443, 2016. K. Olds, R.W. Byard, C. Winskog, N.E. Langlois. Validation of alternate light
CC E
[9]
sources for detection of bruises in non-embalmed and embalmed cadavers. Forensic Sci. Med. Pathol., vol. 13, no. 1, pp. 28-33, 2017.
[10]
D.S. Holbrook, M.C. Jackson. Use of an alternative light source to assess
A
strangulation victims. J Forensic Nurs., vol. 9, no. 3, pp. 140-5, 2013.
[11]
R.M. Limmen, M. Ceelen, U.J.L. Reijnders, J. Stomp et al. Enhancing the Visibility of Injuries with Narrow-Banded Beams of Light within the Visible Light Spectrum. J Forensic Sci., vol. 58, pp. 518-22, 2013.
[12]
M. F. A. M. van Velthoven and H. G. T. Nijs, “Is the visibility of standardized induced bruises improved under a forensic light source?,” in EAFS, Prague, Abstract O211, 2015.